coordinated waves of gene expression during neuronal ... coordinated... · cell types and...

Coordinated waves of gene expression duringneuronal differentiation of embryonic stem cells asbasis for novel approaches to developmentalneurotoxicity testing

B Zimmer1, PB Kuegler1, B Baudis1, A Genewsky1, V Tanavde2, W Koh2, B Tan2, T Waldmann1, S Kadereit1 and M Leist*,1

As neuronal differentiation of embryonic stem cells (ESCs) recapitulates embryonic neurogenesis, disturbances of this processmay model developmental neurotoxicity (DNT). To identify the relevant steps of in vitro neurodevelopment, we implemented adifferentiation protocol yielding neurons with desired electrophysiological properties. Results from focussed transcriptionalprofiling suggested that detection of non-cytotoxic developmental disturbances triggered by toxicants such as retinoic acid (RA)or cyclopamine was possible. Therefore, a broad transcriptional profile of the 20-day differentiation process was obtained.Cluster analysis of expression kinetics, and bioinformatic identification of overrepresented gene ontologies revealed waves ofregulation relevant for DNT testing. We further explored the concept of superimposed waves as descriptor of ordered, butoverlapping biological processes. The initial wave of transcripts indicated reorganization of chromatin and epigenetic changes.Then, a transient upregulation of genes involved in the formation and patterning of neuronal precursors followed.Simultaneously, a long wave of ongoing neuronal differentiation started. This was again superseded towards the end of theprocess by shorter waves of neuronal maturation that yielded information on specification, extracellular matrix formation,disease-associated genes and the generation of glia. Short exposure to lead during the final differentiation phase, disturbedneuronal maturation. Thus, the wave kinetics and the patterns of neuronal specification define the time windows and end pointsfor examination of DNT.Cell Death and Differentiation (2011) 18, 383–395; doi:10.1038/cdd.2010.109; published online 24 September 2010

Ultimately, the entire complexity of the mammalian centralnervous system (CNS) is generated during ontogenesis froma few single cells. Neuronal generation and differentiation canbe recapitulated by embryonic stem cells (ESCs) underappropriate culture conditions.1–6 ESC-based studies ofneurodevelopment allow investigations, which are not easilypossible in vivo.7 However, known differentiation protocolsdiffer in their suitability for toxicological studies. For instance,older protocols involve a step of embryoid body formation.8

Frequently, only a small number of the initially present ESCsform neurons, and the observation of individual cells is hardlypossible. Other protocols use co-cultures with stromal celllines to differentiate ESCs towards neurons, and wouldtherefore introduce additional complexity into models fordevelopmental neurotoxicity (DNT). A recently developedmonolayer differentiation protocol allows monitoring of thedifferentiation procedure and of possible effects of differentchemicals during the whole period of differentiation on a singlecell level.9

DNT is the form of toxicity least examined and hardest totrace, as it is not necessarily related to cell loss. In all, o0.1%

of frequently used industrial chemicals have been examined,and for the few known toxicants, the mechanism of action is stillelusive (reviewed in Makris et al.10; Grandjean and Landrigan11;Bal-Price et al.12). Behavioral pathology in the absence of cellloss is also known from disease models, for example, forHuntington’s disease13 or schizophrenia.14 Toxicants, such asmercury or lead, may trigger behavioral or cognitive deficitswithout histophathological hallmarks.11 Cellular physiology maybe affected during the period of exposure.15 This may eventuallylead to changes in differentiation and patterning in the CNS,which is the basis for long-term effects that are observed afterthe exposure to toxicants has ceased.

CNS development is assumed to be orchestrated by wavesof gene expression16,17 that determine different intermediatecell phenotypes. Some periods may be more sensitive tocertain toxicants than others. Epidemiological proof for such‘windows of sensitivity’ in organ development with long-termconsequences for the organism comes from thalidomideexposure in man3 and various animal models.18

Current test systems based on the differentiation of stemcells to either cardiomyocytes19 or neural cells12 neither yield

Received 06.5.10; revised 14.7.10; accepted 27.7.10; Edited by JC Marine; published online 24.9.10

1Doerenkamp-Zbinden Chair for In Vitro Toxicology and Biomedicine, University of Konstanz, Konstanz D-78457, Germany and 2Bioinformatics Institute, 30 BiopolisStreet, No. 07-01, Singapore 138671, Singapore*Corresponding author: M Leist, Doerenkamp-Zbinden Chair for In Vitro Toxicology and Biomedicine, University of Konstanz, Box 657, Konstanz D-78457, Germany.Tel: þ 49 7531 885037; Fax: þ 49 7531 885039; E-mail: [email protected]: stem cell; development; neurotoxicity; neuron; astrocyte; electrophysiologyAbbreviations: CNS, central nervous system; DNT, developmental neurotoxicity; DoD, day of differentiation; ESC, embryonic stem cells; GO, gene onthology; mESC,murine embryonic stem cell; N, gene onthology neuronal differentiation; NPC, neural precursor cell; RA, all-trans retinoic acid; Shh, sonic hedgehog; GMEM, Glasgow’smodified Eagles medium; FBS, fetal bovine serum; PFA, paraformaldehyde; PBS, phosphate-buffered saline

Cell Death and Differentiation (2011) 18, 383–395& 2011 Macmillan Publishers Limited All rights reserved 1350-9047/11

www.nature.com/cdd

http://dx.doi.org/10.1038/cdd.2010.109

mailto:[email protected]

http://www.nature.com/cdd

mechanistic information, nor do they account for the complex-ity of CNS development, that is, the establishment of abalance between multiple neuronal cell types.3,20 The‘toxicology for the 21st century’ initiative21,22 suggests theidentification of pathways as opposed to the current black-boxtest systems. In the case of ESC-based models of DNT, thisrequires a detailed understanding of the developmentalprocess leading to multiple different cell types. Detailedknowledge on the waves of gene induction controlling differentdevelopmental steps would be an essential prerequisite.However, CNS development is proceeding at different paces.For instance, the anterior and posterior parts of the neuraltube follow different kinetics, and some regions of the CNScontinue neurogenesis, whereas in other regions, cells havealready reached fully post-mitotic stages.20

Our study was undertaken to analyze the wave-likeexpression pattern of mESC neurodevelopment as a basisfor the definition of test windows and markers. This knowledgeshould help to identify non-cytotoxic, but neuroteratogeniccompounds that are able to shift neuronal composition orphenotypes. Finally, the markers should distinguish multiplecell types and differentiation stages, and be able to indicatesubpopulations of cells.

Results

Monolayer differentiation of mESCs to neurons. On dayof differentiation 20 (DoD20), the majority of cells waspositive for the pan-neuronal markers Tuj1 and NeuN. Manycells also expressed the synapse-associated markers SV2and PSD95 (Figure 1a). As a more quantitative overallmeasure for the robustness of the differentiation protocol, wechose mRNA expression, which we followed over time. Thekinetics for different markers was highly reproducibleacross experiments (Figure 1b). Differentiation to mature,electrophysiologically active neurons was shown by thepresence of voltage-dependent Naþ and Kþ , and Ca2þ

channels in individual patch-clamped neurons (Figure 2a–c,Supplementary Figure S1). Further experiments alsoidentified spontaneous neuronal electrical activity(Figure 2d) and action potentials (Supplementary Figure S1).Currents were also evoked by exposure to N-methyl-D-aspartate or kainic acid and blocked by the respectiveselective antagonists (Figure 2e). Thus, our differentiationprotocol yielded bona fide neurons.

Transcription-based end points to identify disturbedneuronal differentiation. We next investigated whethersubtle perturbations of the differentiation process below thecytotoxicity threshold would be detectable by mRNA-basedreadouts. Parallel mESC cultures were differentiated for 7,15 and 20 days and mRNA was prepared for quantitative RT-PCR analysis. These cells were treated during two differenttime windows (DoD1–7, DoD8–15) with two neuroteratogens(Figure 3A). With the concentrations used here, cell deathwas not detectable (data not shown), and cells looked viableand were morphologically indistinguishable from untreatedcells (Figure 3B). We used the morphogen retinoic acid (RA)as a known in vivo and in vitro reproductive toxicant and

cyclopamine for its ability to alter sonic hedgehog (Shh)signaling, resulting in the disruption of patterning gradientsresponsible for floor plate and ventral neurons.2,20 Asexpected from the literature,23 RA induced accele-rated neuronal differentiation (increased synaptophysinexpression) whereas cyclopamine reduced the expressionof markers typical for more ventrally located neurons like

Figure 1 Protein and mRNA-based markers of robust neuronal differentiation ofmESCs. (a) Cultures of mESCs were fixed and stained on day 20 of differentiation.DNA (blue) was stained with H-33342. Proteins are indicated as text on themicrograph in the same color as used for the display of their staining pattern. Tuj1:neuronal form of beta-III tubulin; NeuN: neuron-specific nuclear antigen, encoded byfox3;40 GAD65: glutamate decarboxylase 2; SV2: synaptic vesicle glycoprotein 2a;PSD95: post-synaptic density protein 95. Scale bars: 20 mm. (b) mESC cultures(n¼ 5 biological experiments) were differentiated towards neurons, and RNA wasprepared at the indicated days of differentiation. Gene expression of the stemnessfactor Oct4, the neural stem cell marker Nestin (nes), the mature neuronal markersynaptophysin and the glial marker gfap was quantified by quantitative RT-PCR.The mean±S.D. of the relative expression compared with day 0 (set to 1 on eachdiagram) was calculated and displayed (dotted lines). Relative gene expressiondata were also obtained by chip analysis and the means (n¼ 2) are displayed (solidline). Note the different scaling of the axes for chip or RT-PCR analysis, which waschosen for reasons of better comparability of the overall curve shapes. The figuresin the diagram indicate the relative expression level on DoD20 (DoD7 for nestin)versus DoD0, and thus define the axis scaling

Transcriptional pattern of neuronal differentiationB Zimmer et al

384

Cell Death and Differentiation

Shh, Nkx2.1 and Dlx1, but not overall neuronal differentiation(Figure 3A). Thus, marker genes can indicate subtle shifts indifferentiation patterns not visible morphologically. When

cultures were exposed to cyclopamine from DoD1–7 andimmediately analyzed thereafter, treatment did not affect theoverall formation of neural precursor cells (NPCs), but thereduced Shh expression suggested a reduced ventraldevelopment. In cells left to differentiate further without thecompound, reduced Shh expression was still observable onDoD15. A shift of neurotransmitter phenotype fromGABAergic (Gad2 as marker) to glutamatergic (Vglut1 asmarker)2 was not observed after treatment for the first7 days, but a significant decrease in Gad2 (more ventrallyprominent) was observable when the cells were treatedbetween DoD8 and 15.

In the case of RA, the acceleration of development(synaptophysin) was already significant at early stages andwe found upregulation of markers usually expressed in caudalparts of the neural tube (hoxa6, hb9), and associated with thedevelopment of motor neuron precursors (isl1; Figure 3).

We also examined whether inhibited differentiation wasdetectable by RNA markers. Early exposure to 3i, a kinaseinhibitor mix known for inhibiting differentiation of mESCs,24

resulted in cultures with retarded neural differentiationindicated by a decreased expression of hes5, nestin andTubb3, and an increased expression of Oct4 (SupplementaryFigure S2). Treatment of cells with 3i after DoD7 (after neuraldifferentiation had been initiated) did not return them to thestem cell state, but was cytotoxic. These examples demon-strated the usefulness of transcript profiling for detection ofpatterning disturbances.

It may be necessary to measure the impact on differentia-tion at different DoD, and specific markers need to be selectedfor each stage.

Figure 2 Electrophysiological evidence for successful neuronal development.Cells were differentiated on glass coverslips towards the neuronal lineage for 20–24days and then placed into a temperature-controlled recording chamber for whole-cell patch-clamp studies. (a) Representative example for the currents observedduring the 20 ms voltage steps of the whole-cell voltage clamp recording protocoldisplayed in (b). Note that Naþ currents (downwards deflection) are observed atvoltages X�40 mV (solid line). Strong depolarizing and repolarizing (Kþ currents;upwards deflection) are observed at depolarization to 0 mV (dashed line). (c) Forvoltage clamp recording (voltage step from �80 to 0 mV) of Ca2þ channels, Naþ

and Kþ channels were blocked by addition of tetrodotoxin, tetraethylammo-niumchloride (5 mM), 4-aminopyridine (10 mM) and substitution of intracellular Kþ

ions by 120 mM Csþ . Moreover, the measurement of Ca-currents was favoured bya bath solution containing barium ions (10 mM) instead of calcium ions. Currenttraces were obtained without Ca2þ -channel blocker, or with the blockersnimodipine (1 mM) or Cd2þ (1 mM) added. Current data at 15 ms after the voltagestep were corrected for cell capacitance (indirect measure for cell size) anddisplayed. Data represent means±S.D. **Po0.01. (d) Spontaneous actionpotentials were recorded in current clamp mode (0 pA). At the time indicated byan arrow, tetrodotoxin was added. The dashed line indicates 0 mV membranepotential. The scale bars indicate the dimensions of the membrane potential andthe time domain. (e) Recordings at individual neurons excited with specificglutamate receptor agonists in the presence or absence of blockers. Current traceswere recorded after application of N-methyl-D-aspartate (NMDA) or kainic acid. Allagonists were also tested in the presence of their respective specific antagonist(traces with 5-aminophosphovalerate (AP-5), 6,7-dinitroquinoxalin-2,3-dione(DNQX)). The scale bars represent the current and time dimensions of theexperiment. Data are representative for nX10 neurons (for agonists) and n¼ 3 forantagonists (on neurons with positive agonist response)


385


Identification of clusters of genes regulated duringneuronal differentiation of mESCs. Using oligonucleotidemicrorarrays, we analyzed changes in the transcriptome overtime to identify toxicity markers. The differentiation kineticsidentified in the microarray analysis matched the onesobserved during many other well-controlled experiments(Figure 1b). The kinetics of expression of each generepresented on the chip was used as input for an unbiasedclustering analysis, which yielded eight regulation profiles(Figure 4a, Supplementary Figure S3), besides the genes notregulated at all. Cluster Ia was characterized by rapiddownregulation, and cluster Ib by slow downregulation.These two clusters exemplify the principle of superimposedgene regulation waves with different amplitudes. Clusters IIaand IIb contained genes that were transiently regulated atDoD7 (IIa: up, IIb: downregulated). Cluster IIIa and IIIb werecharacterized by a rapid increase in transcripts between day0 and DoD7, maintained then at high levels. Cluster IVcontained genes, which remained low until DoD7 and thenreached high levels on DoD15. The final cluster V comprisedtranscripts that were hardly upregulated until DoD15, andreached their maximum on DoD20 (Figure 4a).

The genes were subjected to a more detailed analysis. Of40 genes that characterize the initial mESC stage,3 33 wereidentified and all were downregulated (Figure 4b). Most NPCmarkers were found in clusters IIa and IIIa/b, containing genesupregulated early. In contrast to this, most neuronal markerswere found in the clusters with increasing gene expression(III–V), whereas about 20% were found in cluster IIa (transientupregulation on DoD7; Figure 4b). The clusters identified byunbiased bioinformatics methods may therefore correspondto waves of real biological processes underlying the differ-entiation of mESCs to neurons. To explore this workinghypothesis, we continued with an analysis of the biologicalsignificance of genes in individual clusters.

Loss of pluripotency is accompanied by progressivechanges in transcripts responsible for chromatinorganization and DNA/cell cycle functions. Genes incluster I were analyzed for gene onthology (GO) categoriessignificantly overrepresented. Besides the cell cycle, wefound chromatin structure and epigenetic processes to beaffected (Figure 5a, Supplementary Figure S4). All genesknown to be associated with chromosome structure, DNA

Figure 3 Detection of non-cytotoxic developmental disturbances by transcriptional analysis. Cultures of mESCs were neuronally differentiated for 7, 15 or 20 days asindicated in a–d. They were exposed to RA or cyclopamine (Cyclo) for the time periods indicated by the hatched boxes. (A) RNA was isolated at the indicated days (diamond)and used for quantitative RT-PCR analysis of selected differentiation and patterning markers. Headings indicate the overall biological effect, such as accelerated neuronaldifferentiation (e.g. neuronal diff. (þ )) or altered patterning (e.g. caudalization). Names are the official gene names, apart from the following: Vglut1¼ Slc17a7, HB9¼Mnx1.The data indicate relative expression levels in % compared with untreated controls at the same time point and are means±S.D. from two to three independent experiments foreach treatment and exposure schedule. Significance levels (by ANOVA within a given experimental condition) are indicated (*Po0.05, **Po0.01, ***P40.001). Thecomplete data set with S.D. is given in Supplementary Figure S2. (B) Representative images of cultures on DoD15 in condition a. RA- and cyclopamine-treated cultures wereviable, indistinguishable from controls (ctrl.)


386


replication, DNA repair and DNA methylation weredownregulated. Also, most of the genes coding forhistones, histone modifiers, chromatin remodeling andchromatin substructuring were found in clusters Ia/b.Confocal microscopy showed that chromatin distributedrelatively homogeneously over the nucleus in mESCs, butwas organized entirely differently after 20 days ofdifferentiation (Figure 5b).

Correlation of neural precursor formation with a strong,transient change in gene expression levels. We examinedwhether genes of cluster II were specifically linked to theprocess of NPC formation. Nestin was expected, and found,

in cluster IIa. Nestin-positive cells were often arrangedin ring-like structures, reminiscent of rosettes or two-dimensional neural tubes25,26 (Figure 6a). Quantification byflow cytometry analysis confirmed the immunocytochemicalfinding that about 80% of all cells in the culture becamenestin positive (Figure 6b). High synchronization ofdifferentiation was suggested by the sharp expressionprofile of genes in cluster IIa (Figure 6c). Besides nestin,many other genes typically associated with neuroepithelialprecursors (NPCs) and neurogenesis were found in clusterIIa (Supplementary Figure S3). Also some genes associatedwith early, but definitive neuronal development wereidentified (Dll1, Hes3). Cluster IIa also contained apparentlyunspecific genes (e.g. Jak2, Foxd4, Bcl-2, Kif21a, Agtr1a,Moxd1, Aacs, Arl2bp, Scd2). We examined which GOcategories were statistically overrepresented by cluster IIagenes. The GO ‘nervous system development’ emerged witha P-value o10�13, and only neuronal/neurodevelopmentalGOs were identified with the exception of ossification (eightweakly significant genes; Table 1). Thus, genes of cluster IIarepresent an important end point for testing the disturbedproliferation and differentiation during the neuroectodermal/neuronal development time window.

Markers of regional fate decisions in the CNS. Weexamined the expression of regional markers in cluster IIa.Amongst the few markers expressed, those for forebrain(Foxg1) and hindbrain (Hoxa2/b2) were evenly distributed(Figure 6d). Also, in the clusters containing continuouslyupregulated genes (IIIa/b, IVþV), forebrain (Reln), midbrain(En1/2) and hindbrain (Lmx1a or Hoxa1) markers wereevenly distributed (Supplementary Figure S5). Accordingly,our experimental model appears to reflect several parallellines of in vivo neural specification, and the ratios ofexpression of different patterning markers may providesensitive indicators of disturbed neurodevelopment.

Specificity for neuronal induction with respect to glialcells. The transcriptional profile allowed us a detailedanalysis of potentially contaminating non-neuronal cells.Some small GFAP-positive cell areas were reproducibly(1–2 small islands/cm2) identified by immunocytochemistry(Figure 7a). As an unbiased search for overrepresented GOcategories did not result in any hits related to gliogenesis(Table 2), we used a list of 25 astrocyte-related genes3 andfound 11 of them to be upregulated on DoD20 compared withDoD0, with four additional astrocytes-related genestransiently upregulated on DoD7 (Figure 7b). This earlyupregulation of apparent astrocytic markers (e.g. vimentin)may be owing to the generation of radial glia-like NPCs atDoD7. This cell type, as exemplified by the upregulation ofFabp7 in cluster IIIb27 or Ascl1 (¼Mash1),28 shares manymarkers with astrocytes.5 The late upregulation of astroglialmarkers was corroborated by qPCR (Figure 7c). Smallincreases in this astrocytic population may affect toxicitytesting during the later differentiation phases. Microgliaappeared to be absent. The contribution of oligo-dendrocytes appears to be negligible.

Figure 4 Cluster analysis of mRNA time course profiles, and their associationwith distinct phases of differentiation. (a) Gene expression kinetics was determinedfor all genes represented on the chip. An unbiased clustering analysis of the kineticprofiles of all regulated genes was performed. For each cluster (named Ia, Ib, IIa,IIb, IIIa, IIIb, IV, V), the means of the absolute expression level of all genes in therespective cluster, for each analysis time point, are displayed and plotted on alogarithmic scale; n: number of genes in the cluster (b). Number of genes expressedin mESCs (ESCs¼ 40 evaluated, 33 found), NPCs (73 evaluated, 63 found) anddeveloping neurons (N) were analyzed by extensive literature search (mESCs,NPCs) or GO-analysis (N). The relative distribution of these genes across thedifferent clusters was calculated (in %) and displayed (e.g. 65% of all ESC markerswas found in cluster Ia, 35% of all N markers in cluster III)


387


Specificity for neuronal induction with respect to othergerm layer lineages. All GO categories significantlyoverrepresented by the genes of clusters III–V(upregulation on DoD20 versus DoD0) were determinedbioinformatically, and were searched for evidence of non-neural cell type formation. Individual clusters did not indicateany non-neural cell types while representation of neuronalGOs was highly significant (Table 2). Upon pooled analysisof clusters IV and V, the GOs ‘blood vessel development’ and‘muscle organ development’ emerged as significant. Thus, asubpopulation of cells present on DoD20 may display smoothmuscle features.

Waves of clustered genes related to neuronalinduction. For characterization of the cultures, weused non-biased bioinformatics methods to identifyoverrepresented GOs (Table 2). In a complementaryapproach, based on literature and expert judgement,we hand-picked interesting groups of genes (Figure 8).

The major result from this combination of strategies was ourfinding that the differentiation did not proceed as sequence ofsequential steps, but rather involved strongly overlappingprocesses with one underlying large wave (cluster IIIa/b)superseded by shorter waves (cluster IV and V). Forinstance, generation of neurons and axogenesis/growthcone formation seemed to be ongoing in the entire periodfrom DoD7 to DoD20, as indicated by groups ofneuroreceptors and growth cone/axon guidance-relatedgenes in cluster III (Figure 8a and b). A larger group ofgenes associated with synaptic vesicles or the transmissionof nerve impulse only appeared later (cluster IV/V). In thelatest phase, genes associated with ‘responses to stress’ and‘hormonal stimuli’, ‘regulation of extracellular matrixcomponents’ and genes known to be ‘associated withhereditary neurodegenerative diseases’ were stronglyupregulated (Table 2, Figure 8c). Analysis by PCRconfirmed the latter finding. The upregulation for disease-associated genes from DoD0 to DoD20 (n¼ 2

Figure 5 Indication of a progressive change in chromatin organization and epigenetic factors in waves of fast and slow downregulation. Gene lists of relevant processeswere assembled both with the help of the GO database and extensive literature search. The clusters were then queried for the presence of these genes. (a) Processes linked tochromatin or DNA repair and DNA replication are displayed, and for each of them, the number of genes found to be regulated during neuronal differentiation of mESCs isdisplayed in brackets. The individual genes are listed in Supplementary Figure S4. Among the identified genes, four (Smarca1, Myst4, Jmjd3 and Hdac11) are known to beneurospecific and five (Suz12, Ezh2, Bmi1, Cbx2 and Cbx8) are components of the polycomb repressor complexes, which play an important role in differentiation-relatedcontrol of gene promoters. These genes could serve as sensitive markers to detect negative effects of compounds on early developmental processes. For each process, thepercentage of genes present in the different clusters is indicated by colour-coded pie charts. All green shades represent clusters of genes downregulated from DoD0 to DoD20.(b) Changes in chromatin structure during differentiation were visualized by DNA staining with DAPI (green) and confocal microscopic analysis. Left panel: undifferentiatedmESCs; right panels: neuronally differentiated cells on DoD20 that were stained with neuron-specific beta-III tubulin antibody (red). Scale bar: 10 mm


388


differentiations) was: 92- and 16-fold for the Alzheimer’sdisease associated genes App and Mapt, 273-fold for theschizophrenia-associated gene Nrnx1, 91-fold for the prionprotein Prnp and 19-fold/56-fold for the Parkinson’s disease-related genes Pink1/Snca.

We wondered whether toxicants with a purported role in thedevelopmental origins of neurodegenerative diseases (seeSupplementary Figure S6), such as lead, affect this very latephase of neuronal differentiation. The transcript levels of twoneuronal markers and the set of disease-associated geneswere used to examine differences in differentiation. Leadexposure had a dampening effect on the expression of App,Mapt, Nrnx1 and Prnp (Figure 8d). Thus, the knowledge onmarkers together with that of the expected timing of theirexpression provides an ideal toolkit for fine-mapping of subtledevelopmental disturbances.

Discussion

We have here demonstrated a concept of overlapping wavesof gene regulation and suggested its use to define protocols,test windows and end points for DNT testing. Our findingsshould be helpful to close a gap between two highlydeveloped, but isolated disciplines: experimental develop-mental neurobiology and toxicology. The former has beenhighly successful in defining the functional importance,regional expression and cell type association of genes. Thelatter has an urgent need for robust and sensitive markergenes to identify disturbances of development. We showedthat subtle changes in the speed of differentiation, or in dorso–ventral or anterior–posterior patterning owing to toxicants canbe detected by using the right choice of mRNA markers. Suchchanges may be considered in vitro correlates of known

Figure 6 Correlation of neural precursor formation with a transiently upregulated group of genes. (a) On DoD7, cultures were immunostained for the neural stem cellmarker nestin (green) and DNA (red). Scale bar: 100mm. (b) For quantification of nestin-positive NPCs, cells were immunostained for nestin on DoD7 and analyzed by flowcytometry. Data are means±S.D. of seven independent differentiations. ***Po0.001. (c) Relative expression profiles of genes from cluster IIa were calculated bynormalization of expression of each gene to DoD0 expression, which was arbitrarily set to 1. The expression kinetics for each gene within that cluster is displayed. (d) Genesupregulated during neuronal differentiation of mESCs were analyzed for their role in regional specification of the brain and classified accordingly (colour-coding). The numberof genes associated with each of the three chosen subregions of the brain are displayed separately for each regulation cluster. A detailed list of genes with their regionalassignment is given in Supplementary Figure S5


389


teratogenic effects of the chosen compounds. For instance,cyclopamine causes dramatic patterning disturbances (holo-prosencephaly) in a defined period of brain development; RAcauses shifts in the anterior–posterior axis organization,favouring the more posterior parts, as found here by transcriptmarkers. Lead affects multiple neuronal types, which is inagreement with the broad pattern of disturbances found here(see Supplementary Figure S6 for references). The data alsosuggest some warning on the limitations of in vitro–in vivocorrelations. Although our cyclopamine data suggest adisturbance in patterning, they would not indicate a problemin the separation of the forebrain hemispheres, as observed inanimal studies. Thus, observations from stem cell systems willhave a major value for raising alerts on certain compoundsand pinpointing potential mechanisms, while complementarydata from other systems may be required to predict specificeffects on humans.

Transcriptional profiling studies, relying predominantly onbioinformatics analysis, suffer from the weakness and errorsof databases and algorithms. For example, assignment ofgenes to GO categories is not always perfect. For instance,the GO for gliogenesis contains ubiquitous signaling andmetabolic molecules as well as highly specific transcriptionfactors. On the other hand, typical astrocyte markers such asGfap and glutamine synthetase are not members of this GO.Moreover, the equal weight given to ubiquitous versus specificgenes in statistical analysis results in biological skewing. Anadditional problem is the visualization of the large amount ofdata in a form that generates meaningful knowledge. Withthese considerations in mind, we chose to combine bioinfor-matic analysis with classical knowledge-based approaches.During this procedure, the entire hit list of several thousandgenes was manually screened, sorted and annotated. Aconsortium of experts was consulted, and results werecompiled in an open access review format.3 We stronglyadvocate such combined approaches for toxicological sys-tems biology, which is at present driven too strongly bycomputational methods.22,29

Electrophysiology studies have a rather qualitative char-acter, as the cells that were patched may not represent the

entire culture. However, our results fully corroborate earlierfindings that functional neurons can be generated frommESCs.8,30 Immunostaining and quantitative RT-PCR wereused as classical and established methods to link chip-basedtranscript profiling to other experiments that have beenperformed with much higher replicate number. In the future,extensive studies, involving RT-PCR controlled by internalstandards, will be necessary for a quantitative definition of afinal set of markers. Notably, we did not use differences inabsolute values of regulation in the present study as basis forany of our conclusions, and all major conclusions are built ongroups of co-regulated and biologically linked genes asopposed to speculations based on the presence or absenceof a single gene. Even though mRNA correlated well withprotein levels, as for example, in brain inflammation studies,31

our approach should not be interpreted as phenotypedefinition on single cell resolution. The genes grouped withinthe clusters described here are not necessarily expressed inthe same cell and therefore do not automatically describe asingle biological entity. However, with these caveats, we feelthat indicators of disturbances of the default development canbe selected with confidence on the basis of our study.

In the area of developmental toxicology and especially inDNT, cause–effects relationships are still mostly unknown,and human epidemiological data are only available for ahandful of industrial chemicals.11 Rodent data based on theOECD test guideline 426 are available for about 200substances.10 With this lack of human-relevant informationand the better animal database, it appears reasonable for usto perform proof-of-principle experiments for the usefulness ofa new approach in rodent cells first, and to validate humancells against these in case of a positive outcome.

At present, DNT studies are based on, for example,behavioral, cognitive or neuropathological end points, andthe next step towards mechanistic information would be anunderstanding of changes on the level of cells and in geneexpression. The overlapping waves defined here wouldprovide a conceptual framework for this. Such waves (i.e.temporally and spatially shifting activation) of gene expressionare known from many pioneering studies of mammalian

Table 1 GO categories significantly overrepresented in cluster IIa

Biological process (GO)a No. ofgenes in IIa

P-value Examples of upregulatedgenes listed in the GO

Nervous system development 51 3e�14 Neurod4, Nes (nestin), Cdh2 (N-cadherin), Fgf5, Sema5b, Efnb2Regulation of nervous systemdevelopment

17 7e�09 Nefm (neurofilament M), Chrna3 (cholinergic R), Ntrk3, Isl1, Foxg1

Regulation of neurogenesis 16 9e�09 Hoxa2, Smo, Dll1 (delta-like 1), Hes3, Metrn, Ntrk3 (¼Trkc)Neuron projection morphogenesis 13 2e�05 Epha7, Mtap1b, Myh10 (myosin heavy chain), Egr2, Epha7, Isl1Central nervous system development 21 1e�06 Mtap1b (microtubule-associated protein), Bmi, Foxg1, Isl1, Fgfr3Neuron projection regeneration 5 2e�06 Mtap1b, Bcl2, Smo, Chst3 (carbohydrate sulfotransferase)Parasympathetic nervous systemdevelopment

4 4e�06 Hoxb2, Egr2, Smo (smoothened), Hes3 (hairy and enhancer of split)

Neuron development 20 5e�06 Mtab1b, Foxg1, Epha7 (Eph receptor A7), Isl1, Ulk2, Bmpr1bCranial nerve development 5 1e�05 Gli3, Hoxb2, Egr2 (early growth response), Smo, Hes3Dorsal/ventral pattern formation 9 7e�07 SP8, Foxg1, Bmpr1a, Bmpr1b (bone morphogenic protein R.), Hoxa2Tissue development 28 1e�06 Homer1, Prox1 (prospero-related homeobox 1), Fzd2, Sdc1 (syndecan)MAPKKK cascade 12 2e�06 Mapk8, Fgf13, Jak2, Nrg1, Fgfr3, Tgfbr1, Mapk8 (¼ Jnk)Anterior/posterior pattern formation 12 4e�06 Hoxb2, Hoxa2, Tgfbr1 (transforming growth factor, beta receptor)Regulation of ossification 8 3e�05 Smad5, Calca (calcitonin), Sfrp1 (secreted frizzled-rel. protein 1), Egr2

aAll categories are identified by gProfiler bioinformatics analysis, with their P-values indicated after correction by removal of ‘nervous system development’ genes fromnon-neuronal GOs.


390


in vivo CNS development20 and are, for instance, wellcharacterized in high density and resolution in the hippo-campus.32 Waves have also been defined in vitro inmESCs17,33 or differentiating embryonic carcinoma cells.16,34

Here, we extended this concept by relating regulation clustersto underlying biological processes important for toxicitytesting. This translation from developmental biology to the

toxicological perspective defines the windows of sensitivityrelevant for test protocols.

In the field of cardiac development, the mESC-basedembryonic stem cell test has been frequently applied.19

Exposure of cells during the entire test period is confoundedby relatively unspecific toxicity. Therefore, separation ofexposure into the proliferation and differentiation phase hasbeen suggested.35 We want to expand this principle here bysuggesting four relevant test periods. DoD1–7: testing oflineage commitment, efficiency of NPC formation and ofepigenetic changes associated with the transition frompluripotent cells to more committed NPCs. DoD8–15: majorphase of neuronal patterning and vesicle development.DoD15–20: a more unexpected, but highly interesting andrelevant phase, when most proliferation has ceased andmaturation becomes evident by expression of matrix compo-nents, important transporters and disease-associated genes.Our data on lead exposure during this phase show that it willbe of high importance for future testing. DoD20þ has notbeen explored here. It requires further investigation todetermine whether this period can be used as stablereference for neurotoxicity versus DNT, or whether newprocesses such as synaptogenesis, gliogenesis or myelina-tion take a dominant role here.

The major task for the future will be the validation of a largerset of such markers, first with known specific and mechan-istically defined disruptors of developmental pathways, thenwith known DNT compounds, to select the smallest group offinal markers useful for a comprehensive description oftoxicities triggered by the test compounds.

Materials and MethodsMaterials. Unless otherwise mentioned, cell culture media and reagents werefrom Invitrogen (Carlsbad, CA, USA) and accessory reagents from Sigma (Munich,Germany). Antibodies used were anti-Tuj1 (cat. no. MMS-435P; Covance, Munich,Germany), anti-NeuN (cat. no. MAB377; Chemicon, Schwalbach/Ts, Germany),anti-GAD65 (GAD-6; DSHB, Iowa City, IA, USA), anti-SV2 (SV2; DSHB), anti-PSD95 (cat. no. 51-6900; Zymed, Darmstadt, Germany), anti-Nestin (cat. no.MAB353; Chemicon), anti-GFAP (clone: G-A-5; Sigma) and anti-Nestin-647 (clone:25/NESTIN; BD Biosciences, Franklin Lakes, NJ, USA). mRNA primers used werePou5f1-forward: 50-CTCTTTGGAAAGGTGTTCAGCCAGAC-30, Pou5f1-reverse:50-CGGTTCTCAATGCTAGTTCGCTTTCTC-30; Nestin-forward: 50-CTGGAAGGTGGGCAGCAACT-30, Nestin-reverse: 50-ATTAGGCAAGGGGGAAGAGAAGGATG-30;Synaptophysin-forward: 50-GGGTCTTTGCCATCTTCGCCTTTG-30, Synaptophysin-reverse:50-CGAGGAGGAGTAGTCACCAACTAGGA-30; Gfap-forward: 50-GCCCGGCTCGAGGTCGAG-30, Gfap-reverse: 50-GTCTATACGCAGCCAGGTTGTTCTCT-30; Shh-forward: 50-CAGCGGCAGATATGAAGGGAAGATCA-30, Shh-reverse: 50-GTCTTTGCACCTCTGAGTCATCAGC-30; Hes5-forward: 50-CCCAAGGAGAAAAACCGACTGCG-30, Hes5-reverse: 50-CAGCAAAGCCTTCGCCGC-30; Tubb3-forward: 50-GACAACTTTATCTTTGGTCAGAGTGGTGCTG-30, Tubb3-reverse: 50-GATGCGGTCGGGGTACTCC-30; Nkx2.1-forward: 50-TACCACATGACGGCGGCG-30, Nkx2.1-reverse: 50-ATGAAGCGGGAGATGGCGG-30; Dlx1-forward: 50-TCACACAGACGCAGGTCAAGATATGG-30, Dlx1-reverse: 50-AGATGAGGAGTTCGGATTCCAGCC-30; HoxA6-forward: 50-CTGTGCGGGTGCCGTGTA-30, HoxA6-reverse: 50-GCGTTAGCGATCTCGATGCGG-30; Hb9-forward: 50-CGAACCTCTTGGGGAAGTGCC-30, Hb9-reverse: 50-GGAACCAAATCTTCACCTGAGTCTCGG-30; Vglut1-forward:50-GGTCACATACCCTGCTTGCCAT-30, Vglu1-reverse: 50-GCTGCCATAGACATAGAAGACAGAACTCC-30; Gad2-forward: 50-AAGGGGACTACTGGGTTTGAGGC-30, Gad2-reverse: 50-AGGCGGCTCATTCTCTCTTCATTGT-30; Isl1-forward: 50-ACCTTGCGGACCTGCTATGC-30, Isl1-reverse: 50-CCTGGATATTAGTTTTGTCGTTGGGTTGC-30; Tubb3-forward: 50-GACAACTTTATCTTTGGTCAGAGTGGTGCTG-30, Tubb3-reverse: 50-GATGCGGTCGGGGTACTCC-30; Mapt-forward: 50-ACACCCCGAACCAGGAGGA-30, Mapt-reverse:50-GCGTTGGAC GTGCCCTTCT-30; App-forward: 50-TCAGTGAGCCCAGAATCAGCTACG-30, App-reverse: 50-GTCAGCCCAGAACCTGGTCG-30; Pink1-forward: 50-GGGATCTCAAGTCCGACAACATCCT-30, Pink1-reverse: 50-CTGTGGACACCTCAGGGGC-30;

Figure 7 Analysis of glia-associated genes. (a) DoD20 cultures were fixed andstained for GFAP (green; to identify astrocytes) and Tuj1 (red; to identify neurons).The left image shows a representative overview with large neuronal areas and onetypical astrocytic island. The right image shows an astrocytic island in greater detail.Scale bars¼ 100mm. (b) The table indicates the glia-related genes identified in thisstudy, sorted by the cluster of expression kinetics they fell into. Astrocyte-relatedgenes searched for, but not identified here were glutamine synthetase (Glul),S100b, Slc1a2 (Glt-1, Eaat2), Connexin 30/43 (Gjb6/Gja1), NfiA (also foundin oligodendrocytes). Oligodendrocyte-related genes not found here wereATP-binding cassette, sub-family A (Abca2), CNPase (Cnp1), a microtubule-associated protein (Mtap4), myelin-glycoproteins (Omg and Mog), Olig2/3 (Olig2,Olig3), myelin protein zero (Mpz), Ng2 (Cspg4), NfiA. (c) Expression of selectedastrocyte-related genes was monitored by qPCR on day 0, 7, 15 and 20 of twodifferentiations. Data for each differentiation are given individually. The lines indicatethe resulting mean values


391


Snca-forward: 50-ATGGAGTGACAACAGTGGCTGAGA-30, Snca-reverse: 50-CACAGGCATGTCTTCCAGGATTCC-30; Prnp-forward: 50-ACCATCAAGCAGCACACGGTC-30, Prnp-reverse: 50-GACAGGAGGGGAGGAGAAAAGCA-30; Nrnx1-forward: 50-GTGGGGAATGTGAGGCTGGTC-30, Nrnx1-reverse: 50-TCTGTGGTCTGGCTGATGGGT-30; Aqp4-forward:50-GCTCAGAAAACCCCTTACCTGTGG-30, Aqp4-reverse: 50-TTCCATGAACCGTGGTGACTCC-30; Gjb6-forward: 50-CGTACACCAGCAGCATTTTCTTCC-30, Gjb6-reverse: 50-AGTGAACACCGTTTTCTCAGTTGGC-30; SparcL-forward: 50-CCCAGTGACAAGGCTGAAAAACC-30, SparcL-reverse: 50-GTAGATCCAGTGTTAGTGTTCCTTCCG-30; Slc1a3-forward:50-CTCTACGAGGCTTTGGCTGC-30, Slc1a3-reverse: 50-GAGGCGGTCCAGAAACCAGTC-30; Pla2g7-forward: 50-GGGCTCTCAGTGCGATTCTTG-30, Pla2g7-reverse: 50-CAACTCCACATCTGAATCTCTGGTCC-30; Aldh1l1-forward: 50-CTCGGTTTGCTGATGGGGACG-30,Aldh1l1-reverse: 50-GCTTGAATCCTCCAAAAGGTGCGG-30; Pygb-forward: 50-GGACTGTTATGATTGGGGGCAAGG-30, Pygb-reverse: 50-GCCGCTGGGATCACTTTCTCAG-30;Vim-forward: 50-GAGATGGCTCGTCACCTTCGTG-30, Vim-reverse: 50-CCAGGTTAGTTTCTCTCAGGTTCAGG-30. The toxicants used were RA (cat. no. R2625; Sigma), cyclopamine:(cat. no. 239803; Calbiochem, Darmstadt, Germany), PD184352: (cat. no. Axon 1368; AxonMedchem, Groningen, The Netherlands), SU5402 (cat. no. 572631; Calbiochem) andCHIR99021 (cat. no. Axon 1386; Axon Medchem).

Cell culture and differentiation. CGR8, a widely available murine ESC(mESC) line suitable for feeder-free culture maintenance and with establishedpotential to develop along the neuroectodermal and neuronal lineage36,37 waskindly provided by K-H Krause (Geneva). Cells were cultured in complete Glasgow’s

modified Eagles medium (GMEM), supplemented with 10% heat inactivated fetalbovine serum (FBS; PAA, Pasching, Austria), 2 mM glutamax, 100 mM non-essential amino acids, 50mM b-mercaptoethanol, 2 mM sodium pyruvate and1000 U/ml leukemia inhibitory factor (Chemicon). Cells were kept at 37 1C in 5%CO2 on tissue culture plates coated with 0.1% gelatin, and were routinely passagedevery 48 h.

The mESCs were differentiated towards the neural lineage according to theprotocol developed by Ying and Smith.9 At critical steps, we used the followingparameters: cells were plated in the priming phase at 1.2� 105 cells/cm2 incomplete GMEM on 0.1% gelatin-coated Nunclon culture dishes (Nunc,Langenselbold, Germany). Next day, for neural induction, cells were plated ongelatin-coated Nunclon dishes at 104 cells/cm2 in N2/B27 medium (composition asdescribed in Ying and Smith9, for a detailed description of B27 see http://www.paa.com/cell_culture_products/reagents/growthsupplements/neuromix.html).On day 7 of differentiation (DoD7) for neuronal generation and maturation, cellswere replated at 104 cells/cm2 on poly-L-ornithin (10mg/ml) and laminin (10 mg/ml)-coated Nunclon dishes in N2/B27 medium. Cells were fed every other day withcomplete medium change with N2/B27 medium.

Immunostaining and FACS analysis. For immunocytochemical analysis,cells were fixed with methanol (�20 1C) or 4% paraformaldehyde (PFA) inphosphate-buffered saline (PBS) and permeabilized with 0.1% Triton X-100 in PBS.After blocking with 10% FBS, cells were incubated with primary antibodies (Tuj1

Table 2 GO categories that are overrepresented in the clusters comprising genes upregulated during differentiation

Cluster Biologicalprocess (GO)a

No. ofgenes

P-value Examples of upregulated genes

IIIa/b Nervous systemdevelopment

107 2e�32 Hes5, Notch3, Otx1, FoxA2, Nkx2.2, Ntrk3, Nrxn2 (neurexin)

Generation ofneurons

69 6e�23 Sox5, Shh (sonic hedgehog), Wnt3a, Dcx (doublecortin),Nog (noggin)

CNS development 49 1e�16 Zic1, Wnt7a, Fgf8, Pitx2Neurondevelopment

39 7e�13 Gap43, Gprin2 (inducer of neurite outgrowth), App (Abprecursor protein), Reln

Axogenesis 28 3e�12 Cdk5r1 (kinase), EfnB1 (ephrin), Ntng1 (netrin), Stxbp1(syntaxin-binding protein)

Axon guidance 19 3e�10 Apbb1 (APP-binding), Cxcr4, Slit2, Kif5C (kinesin), Ephb1(ephrin-R)

Neuron projection 32 4e�7 Grik5 (glutamate-R), Gria3 (glutamate-R), Cacna1g (Ca2+

channel), Mtap2 (map2)

IV Vesicle 33 2e�7 Sv2a (synaptic vesicle glycoprotein), Syn2 (synapsin),Syt1 (synaptotagmin)

Nervous systemdevelopment

43 4e�7 Neurog2 (neurogenin), Unc5b (netrin-R), Bai2, FoxD1,Egfr, Dner, En1 (engrailed)

V Extracellular matrix 24 10e�11 Dcn (decorin), Col1a1 (collagen), Spon2 (spondin), Lum(lumican), Tnc (tenascin)

Lipid storage 5 4e�6 Apoa1 (apolipoprotein), Gm2a (ganglioside activator), Enpp1,Cav1 (caveolin)

Response to stress 42 1e�5 Hspa2 (heat shock protein), Fas (fas), Fos, Pparg (PPAR-gamma), Pink1, Snca

IV+V Extracellular matrix 39 3e�11 Col1a2 (collagen), Col3a1 (collagen), Ecm1 (extracellularmatrix), Efemp2 (fibulin)

Response tohormone stimulus

38 7e�10 Rbp4 (retinol BP), Rxra, Thra, Rgs9, Igfbp7 (insulin binding)

Nervous systemdevelopment

70 2e�10 Nrxn1 (neurexin), Mapt (tau), Tgfbr2, Dlx1

Blood vesseldevelopment

29 5e�9 Cdh13 (cadherin-H), Prrx1, SphK1 (sphingosine kinase),Cul7 (cullin)

Neuron projection 33 5e�7 Tubb4 (tubulin), Syt1 (synaptotagmin), Psd2, Syt4, Ttyh1(tweety homolog)

Neurogenesis 44 9e�7 Myo6 (myosin), Nrn1 (neuritin), En2 (engrailed), Hoxa1, Lhx5Synaptic vesicles 14 1e�6 Syp (synaptophysin), Slc17a6, Rabac1 (rab acceptor)Muscle organdevelopment

22 3e�6 Gata6, Des (desmin), Myl2 (myosin light chain), Vamp5(vesicle-associated protein)

Transmission ofnerve impulse

23 2e�5 Gria2 (glutamate-R), Slc17a6 (vGlut), Chrnb1 (ACh-R),Kcnmb4 (K+ channel)

aAll categories are identified by gProfiler bioinformatics analysis, with their P-values indicated after correction by removal of ‘nervous system development’ genes fromnon-neuronal GOs.


392


http://www.paa.com/cell_culture_products/reagents/growthsupplements/neuromix.html

http://www.paa.com/cell_culture_products/reagents/growthsupplements/neuromix.html

1 : 1000, NeuN 1 : 200, GAD65 1 : 200, SV2 1 : 200, PSD95 1 : 500, Nestin 1 : 500,Nestin-647 1 : 40, GFAP 1 : 800) overnight. After incubation with appropriatesecondary antibodies, nuclei were counterstained with Hoechst H-33342 dye.Images were taken on the original cell culture dishes using an IX81 invertedmicroscope (Olympus, Hamburg, Germany) equipped with a -40X, NA 0.6 long-range lens and processed using CellP imaging software (Olympus). For confocalmicroscopy, cells were grown on four-well chamber slides (Nunc), fixed with 4%PFA/2% sucrose in PBS and permeabilized with 0.6% Triton X-100 in PBS. Afterblocking with 5% BSA/0.1% Triton X100 in PBS, cells were incubated with Tuj1antibody in blocking buffer for 1 hour at room temperature. After incubation withappropriate secondary antibodies, nuclei were counterstained with DAPI. Confocalimages were taken using a Zeiss LSM 510Meta confocal microscope equipped witha Plan Apochromat -63X, NA 1.4 oil DIC lens. Images were analyzed and processedusing ImageJ.

For flow cytometry, cells were dissociated on DoD7 with accutase, fixed andpermeabilized in Cytofix Buffer followed by Perm Buffer I (both BD Biosciences),and stained with anti-nestin antibody conjugated to Alexa-647, or isotype control.Cells were analyzed with an Accuri C6 flow cytometer (Accuri Cytometers, AnnArbor, MI, USA) and data were processed with CFlow Plus (Accuri Cytometers).

Quantitative PCR and quality control of differentiation. Total RNAof five independent differentiation experiments, performed at different times, withdifferent CGR8 cell batches, and by different operators was isolated at indicatedtime points for marker gene expression analysis using Trizol, the RNA was retro-transcribed with SuperScript II reverse transcriptase, and the resultant cDNAs wereamplified in a Biorad Light Cycler (Biorad, Munchen, Germany) with primers specificfor the genes of interest and designed for a common melting temperature of 60 1C.Real-time quantification for each gene was performed using SybrGreen andexpressed relative to the amount of gapdh mRNA using the 2(�DD C(T)) method.38

For each run, the consistency of conditions and constancy of gapdh amounts in thesamples were controlled by assessment of its absolute cycle number (¼ 18±0.5).

Gene expression analysis. Cells were used for RNA preparation asundifferentiated mESCs before the priming phase (day 0), on DoD7 (beforereplating), DoD15 and DoD20. RNA was extracted from Trizol preparations andpurified using RNeasy Mini prep columns (Qiagen, Hilden, Germany). The total RNAharvested was quantified using a Nanodrop device (Thermo Scientific, Waltham,MA, USA) and its integrity was assessed using Agilent Bioanalyser (Agilent, SantaClara, CA, USA). Illumina TotalPrep RNA Amplification Kit (Ambion, Foster City,CA, USA) and 500 ng total RNA of each sample was used according to themanufacturer’s protocol to produce biotin-labelled cRNAs. For hybridization ontoSentrix Mouse Ref.8 V2 mRNA microarray beadchips (Illumina, San Diego, CA,USA), 750 ng of labeled cRNA were incubated for 16 h at 58 1C. After hybridization,chips were washed, blocked and streptavadin-Cy3 stained. Fluorescence emissionby Cy3 was quantitatively detected using BeadArray Reader Scan. Statisticalanalysis data are based on duplicate samples. Each of the samples containedpooled RNA from two differentiations to further increase robustness ofresults. Technical variation of the chip was minimal as tested by rerun of thesame sample on two different arrays and by comparison of results from twobeadchips within one array.

Data analysis. Original and processed data have been deposited for publicaccess in the EBI Arrayexpress database (accession no. E-TABM-1068, specifiedrelease date; 30 September 2010). For initial processing, data were uploaded toBeadstudio (Illumina) for background subtraction. Further processing (baselinetransformation and normalization to 75 percentile) and analysis was performed withGenespring 9.0 (Agilent), and all normalized expression kinetics data sets wereused as input for an unsupervised non-hierarchical clustering with relation to theaverage of expression of all genes on the chip, using the K-means algorithm. Theeight major clusters were selected for further analysis. Within these, significant geneexpression profiles were selected, based on a minimum regulation of twofold on anyof the time points and on two-way ANOVA, taking into account the regulation rangeand the variation between different arrays.

Patch-clamp recording. For functional characterization, neurons from atleast three independent differentiations were tested for electrophysiological activity.Electrodes with a resistance of 2–5 MO were pulled of borosilicate glass (Clark,G150F; Warner Instruments, Hamden, CT, USA) on a Sutter Instruments (Novato,CA, USA) P-97 horizontal micropipette puller. All experiments were carried out using

Figure 8 Functional assignment of neuronal genes upregulated in differentwaves. A combination of bioinformatics tools and literature information was used tosearch all upregulated clusters for conspicuous biological themes and for genesassociated with them. Themes are displayed and corresponding genes (with originalNCBI gene names) are colour-coded according to the clusters they were found in(displayed graphically besides the legend, with dots on the lines representing DoD0,DoD7, DoD15 and DoD20). (a) Core neurochemical themes. Note a relatively earlyinduction of receptors and channels, compared with late emergence of genescoding for transporters and synaptic vesicles, and those related to neurodegen-erative disease. (b) Themes related to neurite growth indicate an early focus ongrowth cone formation and guidance. (c) Genes related to extracellular matrix aredisplayed. (d) The cells were treated with a non-cytotoxic concentration (assessedby resazurin reduction and LDH release, data not shown) of lead (1 mM) only duringthe last phase of differentiation (DoD14–DoD20). RNA was isolated on DoD20 andused for quantitative RT-PCR analysis of genes associated with neurodevelopmentand known to be involved in neuronal disease. Pink1 and Snca were not affected.Also, their relative increase with respect to the pan-neuronal marker synaptophysinwas not significant. Data indicate relative expression levels in % compared with theuntreated controls on DoD20 and are means±S.D. (n¼ 2). Significance levels(ANOVA) are indicated (*Po0.05, **Po0.01, ***Po0.001)


393


a custom built recording chamber (800ml volume) made of Teflon within atemperature-controlled microscope stage (37 1C). Whole-cell voltage and currentclamp recordings were obtained from cells at DoD20–24. Cells were grown oncoated glass cover slips (10 mm) from DoD7. Whole-cell currents were recordedusing an L/M-EPC-7 amplifier (List Medical Electronic, Darmstadt, Germany),digitised at sampling frequencies between 10–50 kHz using a DigiData 1320AAD/DA converter (Axon Instruments Inc.). The patch pipettes for spontaneous andevoked action potential measurements as well as for the neurotransmitterresponses were filled with (in mM) 90 Kþ -gluconate, 40 KCl, 1 MgCl2, 10 NaCl,10 EGTA, 4 Mg-ATP, 10 HEPES/KOH (pH 7.4 at 37 1C), whereas the bath solutioncontained (in mM) 155 NaCl, 1 CaCl2, 3 KCl, 10 D-(þ )-glucose, 10 HEPES/NaOH(pH 7.4 at 37 1C). The protocol for recording of Naþ and Kþ channels was asfollows: cells were hyperpolarized to �90 mV, and subsequently stepped to adefined voltage as indicated and returned to �70 mV, before the next cycle, with adifferent voltage step, was run. Each cycle took 120 ms. For the neurotransmitterresponse measurements, the different substances were directly added asconcentrated stock solutions to the recording chamber in amounts of 1–10ml.Antagonists were added at least 1 min before the agonists. Recordings wereinitiated within 100 ms after addition of agonists. For the measurement of bariumcurrents through calcium channels, the pipette filling solution contained (in mM) 110CsF, 10 NaCl, 20 TEA-Cl, 10 EGTA, 4 Na2-ATP, 10 HEPES/CsOH (pH 7.4at 37 1C), whereas the bath solution contained (in mM) 130 NaCl, 10 BaCl2,10 D-(þ )-glucose, 5 tetraethylammonium chloride, 10 4-aminopyridine, 0.5tetrodotoxin, 10 HEPES/NaOH (pH 7.4 at 37 1C).

All current signals were normalized against the individual cell capacitances (as asurrogate measure for cell size) and are expressed in current densities (currentdivided by cell capacitance). Liquid junction potentials were measured andcorrected, using the method described by Erwin Neher (1992), except for bariumcurrent measurements. Stimulation, acquisition and data analysis were carried outusing pCLAMP 10.2 (Axon Instruments Inc., Sunnyvale, CA, USA) and ORIGIN 8.0(OriginLab Corp., Northampton, MA, USA). Fast and slow capacitive transientswere canceled online by means of analog circuitry. Residual capacitive and leakagecurrents were removed online by the P/4 method. Series resistance compensationwas set to at least 50%. For analysis, traces were filtered offline at 5 kHz. Cells formeasurements were chosen with respect to their morphological phenotype (small,round, highly elevated (phase-bright) cell bodies with projections of at least fivetimes the cell body diameter, growing in network-like clusters containing at least 20–30 similar cells). The patch pipette was approached to these cells perpendicular tothe plane formed by the cell membrane in the patch region.

Statistics and data mining. The numbers of replicates of each experimentare indicated in figure legends. Data were presented, and statistical differenceswere tested by ANOVA with post-hoc tests as appropriate, using GraphPad Prism4.0 (Graphpad Software, La Jolla, CA, USA). Assignment of significantlyoverrepresented GO categories to different clusters and calculation ofprobabilities of a false-positive assignment was performed by G-profiler (http://biit.cs.ut.ee/gprofiler/39). For coverage of biological domains without appropriateand well-controlled GO category, relevant genes were assembled from the literatureand cross-checked by 2–3 independent specialists. The number of genes withinthese groups identified in this study was indicated in relation to the overall number ofpossible hits or in relation to their distribution over different clusters. The genesdefined in this study as embryonic stem cell markers or neural stem cell (NPC)markers were derived from recent literature.3 Neuronal (N) differentiation markers(n¼ 574) were defined as all members of GO GO:0048699 (generation of neurons)corrected for those genes used as NPC markers. The graphical representation ofidentified genes (or groups) within their biological context is based on the majorgene function as indicated on the NCBI-gene website and the literature. Importantly,members of each identified group were scored according to their suitability asmarkers for a PCR-based quality control of the differentiation pattern in toxicityexperiments. Several selection rounds were run to identify the final set of markersdisplayed as example genes in the tables and some of the figures.

Toxicity experiments. Cells were exposed to chemicals during differentphases of differentiation to test the suitability of the model system for neurotoxicitytesting, and for testing of DNT during defined time windows. RA (1 mM), ‘3i’(a mixture of 0.8mM PD184352, 2 mM SU5402, 3 mM CHIR99021)24 or cyclopamine(1mM) were added to cultures from DoD1–DoD7 or from DoD8–DoD15. Then, theexperiment was ended, or incubation was continued in the absence of chemicals foradditional 6 days. On the final day, RNA was prepared by the Trizol method for PCR

analysis. For morphological observations, the monolayer regions within the culturewells were imaged. Genes were preselected before the analysis as end points forinitial proof-of-concept experiments, and results from all genes chosen arepresented.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgements. The work was supported in part by the Doerenkamp-Zbinden Foundation, the DFG, the EU FP7 project ESNATS (ML, SK), anIRTG1331 fellowship (BZ) and a fellowship from the KoRS-CB (PBK). We aregrateful to Giovanni Galizia and Sabine Kreissl for help with the electrophysiologicalrecordings and indebted to Bettina Schimmelpfennig for invaluable experimentalsupport. The monoclonal antibodies Gad-6 developed by DI Gottlieb and SV2developed by KM Buckley were obtained from the Developmental StudiesHybridoma Bank developed under the auspices of the NICHD and maintained bythe University of Iowa, Department of Biology, Iowa City, IA 52242, USA. We thankKH Krause for the CGR8 mESC-line, and J Vilo and S Ilmjarv for help withbioinformatics analysis.

1. Conti L, Cattaneo E. Neural stem cell systems: physiological players or in vitro entities?Nat Rev Neurosci 2010; 11: 176–187.

2. Gaspard N, Bouschet T, Hourez R, Dimidschstein J, Naeije G, van den Ameele J et al. Anintrinsic mechanism of corticogenesis from embryonic stem cells. Nature 2008; 455: 351–357.

3. Kuegler PB, Zimmer B, Waldmann T, Baudis B, Ilmjarv S, Hescheler J et al. Markers ofmurine embryonic and neural stem cells, neurons and astrocytes: reference points fordevelopmental neurotoxicity testing. Altex 2010; 27: 17–42.

4. Barberi T, Klivenyi P, Calingasan NY, Lee H, Kawamata H, Loonam K et al. Neural subtypespecification of fertilization and nuclear transfer embryonic stem cells and application inparkinsonian mice. Nat Biotech 2003; 21: 1200–1207.

5. Gotz M, Huttner WB. The cell biology of neurogenesis. Nat Rev Mol Cell Biol 2005; 6:777–788.

6. Abranches E, Silva M, Pradier L, Schulz H, Hummel O, Henrique D et al. Neuraldifferentiation of embryonic stem cells in vitro: a road map to neurogenesis in the embryo.PLoS One 2009; 4: e6286.

7. Leist M, Bremer S, Brundin P, Hescheler J, Kirkeby A, Krause KH et al. The biological andethical basis of the use of human embryonic stem cells for in vitro test systems or celltherapy. Altex 2008; 25: 163–190.

8. Strubing C, Ahnert-Hilger G, Shan J, Wiedenmann B, Hescheler J, Wobus AM.Differentiation of pluripotent embryonic stem cells into the neuronal lineage in vitro givesrise to mature inhibitory and excitatory neurons. Mech Dev 1995; 53: 275–287.

9. Ying QL, Smith AG. Defined conditions for neural commitment and differentiation. MethodsEnzymol 2003; 365: 327–341.

10. Makris SL, Raffaele K, Allen S, Bowers WJ, Hass U, Alleva E et al. A retrospectiveperformance assessment of the developmental neurotoxicity study in support of OECD testguideline 426. Environ Health Perspect 2009; 117: 17–25.

11. Grandjean P, Landrigan PJ. Developmental neurotoxicity of industrial chemicals. Lancet2006; 368: 2167–2178.

12. Bal-Price AK, Hogberg HT, Buzanska L, Lenas P, van Vliet E, Hartung T. In vitrodevelopmental neurotoxicity (DNT) testing: relevant models and endpoints.Neurotoxicology 2009 (in press).

13. Hansson O, Petersen A, Leist M, Nicotera P, Castilho RF, Brundin P. Transgenic miceexpressing a Huntington’s disease mutation are resistant to quinolinic acid-induced striatalexcitotoxicity. Proc Natl Acad Sci USA 1999; 96: 8727–8732.

14. Penschuck S, Flagstad P, Didriksen M, Leist M, Michael-Titus AT. Decrease inparvalbumin-expressing neurons in the hippocampus and increased phencyclidine-induced locomotor activity in the rat methylazoxymethanol (MAM) model of schizophrenia.Eur J Neurosci 2006; 23: 279–284.

15. Rossi AD, Larsson O, Manzo L, Orrenius S, Vahter M, Berggren PO et al. Modifications ofCa2+ signaling by inorganic mercury in PC12 cells. FASEB J 1993; 7: 1507–1514.

16. Wei Y, Harris T, Childs G. Global gene expression patterns during neural differentiation ofP19 embryonic carcinoma cells. Differentiation 2002; 70: 204–219.

17. Aiba K, Sharov AA, Carter MG, Foroni C, Vescovi AL, Ko MS. Defining a developmentalpath to neural fate by global expression profiling of mouse embryonic stem cells and adultneural stem/progenitor cells. Stem Cells 2006; 24: 889–895.

18. Jongen-Relo AL, Leng A, Luber M, Pothuizen HH, Weber L, Feldon J. The prenatalmethylazoxymethanol acetate treatment: a neurodevelopmental animal model forschizophrenia? Behav Brain Res 2004; 149: 159–181.

19. Marx-Stoelting P, Adriaens E, Ahr HJ, Bremer S, Garthoff B, Gelbke HP et al. A review ofthe implementation of the embryonic stem cell test (EST). The report and


394


http://biit.cs.ut.ee/gprofiler/

http://biit.cs.ut.ee/gprofiler/

recommendations of an ECVAM/ReProTect Workshop. Altern Lab Anim 2009; 37:313–328.

20. Rao MS, Jacobson M. Dev Neurobiol. Kluwer Academic/Plenum Publishers: New York, 2005.21. Collins FS, Gray GM, Bucher JR. Toxicology: transforming environmental health

protection. Science 2008; 319: 906–907.22. Leist M, Hartung T, Nicotera P. The dawning of a new age of toxicology. Altex 2008; 25:

103–114.23. Irioka T, Watanabe K, Mizusawa H, Mizuseki K, Sasai Y. Distinct effects of caudalizing

factors on regional specification of embryonic stem cell-derived neural precursors.Brain Res Dev Brain Res 2005; 154: 63–70.

24. Ying QL, Wray J, Nichols J, Batlle-Morera L, Doble B, Woodgett J et al. The ground state ofembryonic stem cell self-renewal. Nature 2008; 453: 519–523.

25. Elkabetz Y, Panagiotakos G, Al Shamy G, Socci ND, Tabar V, Studer L. Human ES cell-derived neural rosettes reveal a functionally distinct early neural stem cell stage.Genes Dev 2008; 22: 152–165.

26. Zhang SC, Wernig M, Duncan ID, Brustle O, Thomson JA. In vitro differentiation oftransplantable neural precursors from human embryonic stem cells. Nat Biotechnol 2001;19: 1129–1133.

27. Feng L, Hatten ME, Heintz N. Brain lipid-binding protein (BLBP): a novel signaling systemin the developing mammalian CNS. Neuron 1994; 12: 895–908.

28. Battiste J, Helms AW, Kim EJ, Savage TK, Lagace DC, Mandyam CD et al. Ascl1 definessequentially generated lineage-restricted neuronal and oligodendrocyte precursor cells inthe spinal cord. Development 2007; 134: 285–293.

29. Hartung T, Hoffmann S. Food for thought y on in silico methods in toxicology. Altex 2009;26: 155–166.

30. Lee S-H, Lumelsky N, Studer L, Auerbach JM, McKay RD. Efficient generation of midbrainand hindbrain neurons from mouse embryonic stem cells. Nat Biotech 2000; 18: 675–679.

31. Lund S, Christensen KV, Hedtjarn M, Mortensen AL, Hagberg H, Falsig J et al. Thedynamics of the LPS triggered inflammatory response of murine microglia under differentculture and in vivo conditions. J Neuroimmunol 2006; 180: 71–87.

32. Mody M, Cao Y, Cui Z, Tay KY, Shyong A, Shimizu E et al. Genome-wide gene expressionprofiles of the developing mouse hippocampus. Proc Natl Acad Sci USA 2001; 98:8862–8867.

33. Schulz H, Kolde R, Adler P, Aksoy I, Anastassiadis K, Bader M et al. The FunGenESdatabase: a genomics resource for mouse embryonic stem cell differentiation. PLoS One2009; 4: e6804.

34. Przyborski SA, Smith S, Wood A. Transcriptional profiling of neuronal differentiation byhuman embryonal carcinoma stem cells in vitro. Stem Cells 2003; 21: 459–471.

35. van Dartel DA, Zeijen NJ, de la Fonteyne LJ, van Schooten FJ, Piersma AH. Disentanglingcellular proliferation and differentiation in the embryonic stem cell test, and its impact on theexperimental protocol. Reprod Toxicol 2009; 28: 254–261.

36. Suter DM, Tirefort D, Julien S, Krause KH. A Sox1 to Pax6 switch drives neuroectoderm toradial glia progression during differentiation of mouse embryonic stem cells. Stem Cells2009; 27: 49–58.

37. Conti L, Pollard SM, Gorba T, Reitano E, Toselli M, Biella G et al. Niche-independentsymmetrical self-renewal of a mammalian tissue stem cell. PLoS Biol 2005; 3: e283.

38. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-timequantitative PCR and the 2-[Delta][Delta]CT method. Methods 2001; 25: 402–408.

39. Reimand J, Kull M, Peterson H, Hansen J, Vilo J. g:Profiler – a web-based toolset forfunctional profiling of gene lists from large-scale experiments. Nucl Acids Res 2007; 35:W193–W200.

40. Kim KK, Adelstein RS, Kawamoto S. Identification of neuronal nuclei (NeuN) asFox-3, a new member of the Fox-1 gene family of splicing factors. J Biol Chem 2009; 284:31052–31061.

Supplementary Information accompanies the paper on Cell Death and Differentiation website (http://www.nature.com/cdd)


395


http://www.nature.com/cdd

coordinated waves of gene expression during neuronal ... coordinated... · cell types and...

Documents